Data Scientist

Role: Data Scientist

Experience: 7-12years

Mode of work: Hybrid

Job Summary

We are looking for an experienced Data Scientist Generative AI with strong expertise in Google Vertex AI, Retrieval Augmented Generation (RAG), Prompt Engineering, Large Language Models (LLMs), and OpenAI technologies. The candidate will design, develop, and deploy AI-powered solutions that leverage enterprise data, advanced NLP techniques, and cloud-based AI platforms.

Key Responsibilities

Generative AI & LLM Development

• Design and develop Generative AI solutions using LLMs, foundation models, and multimodal AI capabilities.
• Work with models from platforms such as OpenAI, Google Gemini, Vertex AI Model Garden, and open-source LLMs.
• Build AI assistants, knowledge management systems, document intelligence solutions, and conversational AI applications.
• Evaluate and fine-tune LLM performance for enterprise use cases.

Vertex AI & Google Cloud AI

• Develop and deploy ML/GenAI solutions using Google Cloud Vertex AI.
• Build pipelines using Vertex AI Studio, Vertex AI Workbench, Model Registry, and Generative AI APIs.
• Implement model deployment, monitoring, evaluation, and governance workflows.
• Integrate Vertex AI with enterprise data sources and cloud services.

Retrieval Augmented Generation (RAG)

• Design and implement scalable RAG architectures.
• Build document ingestion pipelines including:
• Data extraction
• Chunking strategies
• Embedding generation
• Vector search
• Retrieval optimization
• Work with vector databases such as:
• Vertex AI Vector Search
• Pinecone
• FAISS
• ChromaDB
• Weaviate
• Improve retrieval accuracy using hybrid search, reranking, metadata filtering, and evaluation frameworks.

Prompt Engineering

• Create and optimize prompts for LLM-based applications.
• Develop:
• Few-shot prompting
• Chain-of-thought prompting strategies
• Structured output prompting
• Agent-based workflows
• Prompt templates
• Perform prompt evaluation and optimization for accuracy, relevance, and cost efficiency.